Attention and hyperactivity problems are of concern to parents, teachers, clinicians, and researchers. While the familiar transmission of the clinical expression of these problems, namely ADHD (attention-deficit hyperactivity disorder), has been acknowledged, less is known about differences with respect to its components, inattention and hyperactivity/impulsivity. Genetically informative data sets will be analyzed with latent class analysis and structural equation modeling to disentangle the evidence of familial transmission into its genetic and environmental components, and enable us to test for differences in etiology for the two components. We propose to analyze teacher, parent, and self ratings for children and adolescents from three unique data sets. These data sets are collected in studies of (1) adopted and non-adopted probands and their siblings, (2) same-sex twin pairs, and (3) same- and opposite sex twin pairs. Attrition is generally low; structural equation modeling methods will be applied which are capable of dealing with missing data. Apart from being genetically informative, these samples provide additional valuable information due to their longitudinal, rather than cross-sectional, data collection. Combining these data provides powerful tests of sex differences, stability and change over time, as well as associations with covariates (other behaviors, cognitive and family environment data). Preliminary studies have shown different patterns of association of inattention and hyperactivity with other aspects of problem behavior and with cognitive measures. Results so far have been inconclusive, partly due to small sample sizes. Knowledge of differences in genetic and environmental to aspects of attention problems and co-occurring behavior problems will contribute to prevention and intervention planning.